Preamp Shaper Update Mitch Newcomer Andrew Townley Prepared for Munich Liquid Argon Week 2011 Status April 2011 Installed IHPs Cadence IC6 design tools Some issues identified with PDK but mostly OK ID: 371026
Download Presentation The PPT/PDF document "Barrel EM Calorimeter" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Barrel EM Calorimeter Preamp / Shaper Update
Mitch Newcomer, Andrew
Townley
Prepared for Munich Liquid Argon
Week 2011Slide2
Status April 2011
Installed IHP’s Cadence IC6 design tools. Some issues identified with PDK but mostly OK.
A Preamp Design and layout is near completion in IHP’s SG25H3P ( Complementary Bipolar
process ).Alternative Preamp configurations are being considered for layout. Discussion underway with IHP to collaborate on measurements of their PNP devices.
2
Liquid Argon Week Munich 2011Slide3
3
Design Overview & Constraints
Preamp constraints:
Wide input dynamic range
~70nA – 5mA (16 bit)Accuracy target 13 bits
Present fixed input impedance (25
Ω
) across full range
Linear response across input rangeVariable detector capacitance50pF – 1nFAble to drive 120 Ω resistance in shaper stage
C
det
I
in
Preamp
Multi Gain Shaper
Gain Selector
Typical input current waveform
3
Liquid Argon Week Munich 2011
X1,10,10
0Slide4
Preamplifier
Shaping depends on detector capacitance
Increased
Cdet
:
Increases
t
peak
Reduces output amplitude.
Increases series noise
Net: Increased capacitance -> worse SNR
Cdet
= 50pF
Cdet
= 200pF
Cdet
= 1nF
Input current in above: 5mA peak
4
4
Liquid Argon Week Munich 2011Slide5
Preamplifier Circuit
Total Preamp Power ~ 45mW
Ideal Shaping elements
5
Liquid Argon Week Munich 2011Slide6
Input Stage
Q1: minimize series noise
Resistor feedback:
Prevent interconnect
parasitics
from increasing 1.25
Ω
.
Large transient feedback currents ~100mA.
Input feedback transistor – noise critical!
1mA
V
CE
= 3.6V
0.7mA
9mA
V
CE
=
1.5V
V
CE
=
1.5V
V
CC2
=
5V
V
CC1
=
2.5V6
Liquid Argon Week Munich 2011Slide7
Input Stage
vgain
vfollow
I1
I(Rgain1)
V
CE
(Q2)
V
CE
(Q1)
50pF
7
Liquid Argon Week Munich 2011Slide8
Input Stage
Layout considerations:
How to connect to feedback while having minimal impact on resistance ratio?
Low impedance connection to input pad
2R
F1
2R
F1
R
F2
8
Liquid Argon Week Munich 2011Slide9
Feedback resistors
Used lowest per-square resistance available
Easiest way to achieve 1.25
Ω
resistorLess than one square of resistance
Tradeoff: ~30
Ω
feedback resistor ends up being very large
Will be some uncertainty in 1.25ΩSolution: only include 20Ω out of 30Ω on chipExternally tunableSplit 20Ω into two parallel 40Ω
Avoid current crowding effects
600µm
100µm
360µm
60µm
9
Liquid Argon Week Munich 2011Slide10
Input Stage
Input transistor
Q1
NPN_MV:
Nx
=8,
Ny
=1, m=8
12µm
Emitter connection: L=150µm
TopMetal1
Emitter connection: W=400µm
10
Liquid Argon Week Munich 2011Slide11
Feedback resistors
RF1, RF2 set input impedance
Need to dissipate ~100mA peak
Also must be of same type (for matching)Width determined by power density allowance
Max density in
Rsheet
larger than allowed by contact density
Possible to make
addt’l
contact row??
0.75 µm
(1 cont. per 0.75µm) × (0.4mA per contact) =
0.53mA/µm effective max density
Self Heating
I (
mA
)
Resistance
Safe Area
11
Liquid Argon Week Munich 2011Slide12
Output driver
Wide ground, output connections for low impedance.
Minimize potential for current crowding.
TopMetal1 (Out)
2R
BQ1
2R
BQ1
400µm
TopMetal1 (Out)
TopMetal2 (GND)
PNP current mirror
Q
4
Q
4
Q
5
(distributed)
12
Liquid Argon Week Munich 2011Slide13
Output driver
Output driver block
PNP “inverts” voltage signal as current
RC2 converts to voltage
Q5 draws more current (connected to feedback point)
Same function PNP or PMOS?
PNP
Vceo
Limit (2.5V)PNP Vcbo (4V)
13
Vce
= 3.7V
Vce =
3.7V
driverIn
vc_pnp
out
13
Liquid Argon Week Munich 2011Slide14
Output driver
PMOS, not PNP?
Higher
Vbreakdown. (3.3V)Output impedance set by
gm of PMOSReduces effective overall gain when driving low-impedance of feedback
14
Liquid Argon Week Munich 2011Slide15
Output driver
PNP:
Output driver gain 0.96
PMOS:
Output driver gain 0.73. ( First try)
15
Liquid Argon Week Munich 2011Slide16
Chip level
Target: Two or four preamps per chip:
PMOS and PNP?
Add test structures in extra space.Look at breakdown, noise from different transistors.
16
Liquid Argon Week Munich 2011Slide17
Summary and Plans
IHP SG25H3P process is relatively expensive and may offer significant advantages.
PNP
Vceo of 2.5V vs operating point of 3.2 to 3.7V
is the only potential issue identified with the process. Transistor is in a safe operating point but the concern is that spontaneous breakdown currents may occur adding to the amplifier noise. Plan to submit one or two versions of the preamp in the July 2011 run to understand this issue.
IHP has expressed interest in working in collaborating on measurements of the
radiation sensitivity
of the IHP
PNP transistors. Details are under discussion. 17Liquid Argon Week Munich 2011